2019 Volume 44 Issue 12 Pages 849-857
Abuse of recreational drugs (i.e., synthetic chemicals with the structure or expected neurotropic effects, or both, similar to those of controlled substances) is a serious and continuous social harm. Designer drugs are often manufactured or synthesized in small-scale clandestine laboratories with impure starting materials, poor handling skills and inferior storage conditions. Therefore, in addition to the objective compound, diverse impurities may be present, for example, from the starting material, intermediates, catalytic metals formed during chemical synthesis, and materials from the environment. Impurity profiling of drug seizures is a useful scientific tool to obtain information on the clandestine manufacturers and drug trafficking networks. 1-Phenyl-2-(1-pyrrolidinyl)-1-pentanone (α-PVP), a novel psychoactive substance of the cathinone type that is banned in many countries, is still supplied and distributed within the illicit drug market. By using GC-MS and ICP-MS, we identified and estimated the relative contents of organic and inorganic impurities in the bulk powder of 15 batches of α-PVP. We then conducted multivariate data analyses to reveal characteristic patterns of the profiles. Hierarchical cluster analysis of both the organic and inorganic impurities revealed two groups that showed similar impurity profiles, which suggested that the batches in these groups were synthesized in similar routes under similar synthetic environments. The initial groups revealed by the organic impurities were further divided when combined with the data from the inorganic impurities. The present study, therefore, demonstrated the effectiveness of integrated analyses of organic and inorganic impurities for the accurate clustering of designer drugs, to provide precise information to drug investigation authorities.
Abuse of recreational drugs remains a serious and continual social harm. Recreational drugs are synthetic chemicals with similar structures or expected neurotropic effects, or both, of controlled substances. These chemicals can, therefore, illicit unexpected reactions in abusers in addition to the effects of known stimulants and narcotics. In addition, designer drugs and their ingredients are often manufactured or synthesized in small clandestine laboratories with impure starting materials, poor handling skills and inferior storage conditions. Therefore, diverse impurities are often present beside the objective compounds. Such impurities may include chemical precursors, intermediates, reaction byproducts, degradation products, catalytic metals and materials from the environment. Impurity profiling of drug seizures can help to obtain information on the clandestine manufacturers and drug trafficking networks. For example, organic impurities in illicit methamphetamine and related compounds have been reported from different countries and areas (Kishi et al., 1983; Raihana et al., 2015; Zhang et al., 2008). However, information regarding impurities of designer drugs still remains limited.
1-Phenyl-2-(1-pyrrolidinyl)-1-pentanone (α-pyrrolidinopentiophenone or α-PVP) is a novel psychoactive substance of the cathinone type (Fig. 1) with a chemical structure that resembles cathinone, an alkaloid obtained from Khat (Catha edulis; Watterson and Olive, 2014). α-PVP acts on dopamine and noradrenaline transporters, causing hallucinations, arousals, aggression and euphoria (Hataoka et al., 2017; Kaizaki et al., 2014; Kolesnikova et al., 2019; Nóbrega and Dinis-Oliveira, 2018). Currently, α-PVP is banned in many countries, but it is still supplied to and distributed within the illicit drug market, which has resulted in a serious societal issue (Lirio, 2014). We obtained different batches of α-PVP during a market survey that was conducted before the substance was regulated. We identified and estimated the relative contents of organic and inorganic impurities in the bulk powder of α-PVP using GC-MS and ICP-MS. We then analyzed the findings using a multivariate data analysis to reveal characteristic profile patterns.
Chemical structures of 1-phenyl-2-(1-pyrrolidinyl)-1-pentanone (α-PVP).
Fifteen batches of α-PVP were obtained from the street-drug market before it was regulated as a designated drug in Japan in November 2012. α-PVP was handled under the professional supervision of a narcotic researcher (S. Numazawa) licensed under the Narcotics and Psychotropic Control Act of Japan. The signals and composition, as determined by GC-MS, element analysis, and NMR, coincided well with those of authentic α-PVP, as previously reported (Kaizaki et al., 2014), and the purity was considered to be 75% or more.
Sample preparationFor the GC-MS analysis, 10 mg of the bulk powder of each α-PVP sample was dissolved in methanol to make a 1 mg/mL solution. For the ICP-MS analysis, 10 mg of the bulk powder of each α-PVP sample was dissolved in 1% Ultrapur nitrate (Kanto Chemical, Tokyo, Japan) to make a 5 mg/mL solution. For calibration, we used ICP mixed standard solution F (Ce, Dy, Er, Eu, Gd, Ho, La, Lu, Nd, Pr, Sm, Sc, Tb, Tm, Yb and Y), ICP mixed standard solution G (Au, Hf, Ir, Pd, Pt, Rh, Ru, Sb, Sn and Te), ICP mixed standard solution H (Al, As, Ba, Be, Bi, Ca, Cd, Co, Cr, Cs, Cu, Fe, Ga, K, Li, Mg, Mn, Na, Ni, Rb, Pb, Se, Sr, V and Zn), and standard stock solutions of Ag, Te and Tl that were obtained from Kanto Chemical. The Hg standard solution was purchased from Fujifilm Wako Pure Chemical Corporation (Osaka, Japan). α-PVP solution was diluted with 1% Ultrapur nitric acid to prepare a set of calibration standards (0.01, 0.1, 1, 10 and 100 ppb). Plastic tubes used for sample preparation were immersed in 5% Ultrapur nitric acid overnight and washed with ultrapure water.
GC/MS analysisOrganic impurities were analyzed using a GCMS-QP2010 Ultra with AOC-20i as an auto injector (Shimadzu Co., Kyoto, Japan). The analytes were separated on a DB-5MS capillary column (30 m × 0.25 mm I.D., 0.25 mm film thickness; Agilent J & W, Folsom, CA, USA). The splitless injection mode was used. The vaporization chamber temperature was set at 260°C. The initial oven temperature of 60°C was held for 2 min, followed by a 10°C/min increase to 320°C which was held for 12 min; resulting in a total analysis time of 40 min. Data were collected using a scan mode (35–700 m/z). For the identification of chemicals, library searches of mass spectra were performed using the SWGDRUG MS Library ver. 1.7, GC/MS Forensic Toxicology Database (Shimadzu Co.) and NIST mass spectral library.
ICP-MS analysisInorganic impurities were analysed using iCAPQc ICP-MS (Thermo Fisher Scientific, Waltham, MA, USA). The operating parameters were as follows: RF power, 1.55 kW; nebulizer flow, 1.04 L/min; auxiliary flow (argon), 0.8 L/min; and cool gas flow rate, 14.0 L/min. The kinetic energy discrimination mode was used to measure all elements. The quantification limit for Se and Hg, and for Fe, As and Te in the α-PVP bulk powders were ≥ 2 ppb and 0.2–2.0 ppb, respectively. Other elements could be detected with quantification limits as low as 0.2 ppb.
Multivariate analysisMultivariate analyses were conducted using JMP 13 Pro (SAS Institute Inc., Cary, NC, USA). Data were analyzed using a principal component analysis (PCA) and hierarchical cluster analysis (HCA). The impurity dendrogram of the α-PVP batches was constructed using a HCA based on Ward method (Anderberg, 1973).
The organic impurities contained in 15 batches of α-PVP bulk powder were analyzed by GC-MS. In total, 34 components other than α-PVP were tentatively identified by their fragmentation patterns compared with the compound databases (Table 1). Of the 15 batches, 7 contained methylenedioxypyrovalerone (MDPV), 4 contained α-pyrrolidinobutiophenone (α-PBP), 2 contained α-pyrrolidinooctanophenone (PV9), and pentedrone, pentylone, α-pyrrolidinopropiophenone (α-PPP), 4’-methyl-α-pyrrolidinopropiophenone (4-MePPP) and N-propyl-3,4-methylenedioxyamphetamine were observed in a single batch. All of these cathinone- and amphetamine- type compounds became regulated in Japan in 2012 or thereafter.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/44_849-t1.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
To roughly estimate the contents of these chemicals, we calculated their peak area ratios to α-PVP. Although three batches were found to contain relatively high amounts of MDPV (0.88–1.33%), the other batches contained less than 0.1% of this chemical. Therefore, these drugs may have unintentionally been mixed into the α-PVP bulk powder. It is also possible that the production line was shared with multiple drugs. Valerophenone was found in all batches, which suggests that it was used as a starting material for the synthesis of α-PVP (Fig. 2).
Chemical synthesis of α-PVP.
The relative contents of the 34 kinds of organic impurities, which were subjected to HCA based on Ward’s method, were used to construct the dendrogram (Fig. 3). Ward’s method is used to create groups, based on the minimized variance within the clusters. The HCA based on the organic impurities indicated that 15 batches of α-PVP could be divided into 3 groups: group A, comprised of batches No. 1 and 6; group B, comprised of batches No. 2–5 and No. 7–9; and group C, comprised of batches No. 10–15. To investigate factors involved in the grouping, the substances and contents of the 34 organic impurities in the 15 batches were analyzed using PCA, i.e., a multivariate analysis that indicates the contribution of variables to a particular grouping (Fig. 4). Consistent with the HCA, the PCA score plot also indicated that the 15 batches of α-PVP could be divided into three groups. The PCA loading plot indicated the contributions of the organic impurities to the groupings. In group A, (S)-2-[N’-(N-benzylprolyl)amino] benzophenone, 2-bromo-1-phenyl-1-propanone, benzoic acid, glyoxylic acid, N-benzoyl-L-proline and pyrrolidine showed high contribution ratios. In group B, 2-pentanone, 3-pyridenamide, benzamide, diethyl phthalate, MDPV, N-propyl-3, 4-methylenedioxyamphetamine and α-PBP showed high contribution ratios, which indicated that this group was highly contaminated with cathinone type drugs, in addition to α-PVP. In group C, 3-(methylsulfanyl)propanenitrile, dimethyl sulfone, S-methyl methanethiosulphonate and valerophenone showed high contribution ratios; especially that of dimethyl sulfone (Table 1), a highly polar solvent that is industrially used during the chemical synthesis and is often detected in crystalline methamphetamine seizures. It is suggested that dimethyl sulfone was mixed intentionally by the manufacturer as an adulterant agent in the production of group C α-PVP bulk powder (Inoue et al., 2008). In addition, 3-(methylsulfanyl) propanenitrile is used in the synthesis route following the Grignard reaction (Casale and Hays, 2012), which suggests that the α-PVP in group C was synthesized using this method.
Cluster dendrogram of the α-PVP batches based on the 34 detected organic impurities. Numbers represent the batch number. The broken line indicates the division of the 15 batches of α-PVP into 3 groups (groups A, B and C).
Principal component analysis of the 15 batches of α-PVP based on the 34 organic impurities that were identified by GC-MS. The score plot (i) and loading plot (ii) are indicated.
In this study, GC-MS analysis was conducted without derivatization to prevent contamination from derivative reagents. Therefore, it is possible that a trace amount of compounds or highly polar components were present in addition to the listed chemicals. It may, therefore, be necessary to modify the analytical conditions, e.g., using LC-MS/MS, to analyze these compounds. In addition, the relative contents of chemicals were determined as peak area ratios to α-PVP in the present study and, therefore, the findings may not exactly mirror the actual contents. Precise analyses, including quantitation, should be conducted using authentic standards in future studies.
The inorganic impurities contained in the 15 batches of α-PVP bulk powder were determined following another series of analyses by ICP-MS. A total of 37 elements were identified in sufficiently higher contents than the limits of quantification (Table 2). These data were subjected to HCA (Fig. 5), which indicated that the 15 batches of α-PVP could be divided into 3 groups: group D, comprised of batched No. 1, 2, 14 and 15; group E, comprised of batches No. 5, 6 and 9–13; and group F, comprised of batches No. 3, 4, 7 and 8. To investigate the factors involved in the grouping, the 37 elements and their contents in the 15 batches of α-PVP were analyzed by PCA (Fig. 6). Consistent with the HCA, the PCA score plot also indicated that the 15 batches of α-PVP could be divided into 3 groups. The PCA loading plot indicated that group D had relatively high contribution ratios of the noble metals, such as Ag, Au, Ir and Pt. Group E was characterized by less distinctive features of elements. In group F, Al, As, Cr, Cu, Ga, Mn, Ni and Pb showed high contribution ratios.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/44_849-t2.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Cluster dendrogram of the α-PVP batches based on the 37 detected inorganic impurities. Numbers represent the batch number. The broken line indicates the division of the 15 batches of α-PVP into 3 groups (groups D, E and F).
Principal component analysis of the 15 batches of α-PVP based on the 37 kinds of elements identified by ICP-MS. The score plot (i) and loading plot (ii) are indicated.
Data of the organic and inorganic impurities were combined and subjected to HCA, based on Ward’s method, to construct the dendrogram (Fig. 7). The combined organic and inorganic impurities data indicated the presence of two distinctive groups that were comprised of No. 10–13 and No. 3, 4, 7 and 8, respectively. These groups showed similar profiles in the HCA, which suggested that the batches in these two groups were synthesized in similar routes under similar synthetic environments. The integrated analysis of organic and inorganic impurities was found to be useful for the clustering of the batches, because the groups based on the organic impurities were able to be further divided when combined with the data from the inorganic impurities. Therefore, the present study demonstrated the usefulness of integrated analyses in the accurate clustering of designer drugs, to provide precise information to drug investigation authorities.
Hierarchical cluster analysis of the 15 batches of α-PVP based on 34 organic and 37 inorganic impurities. The broken line indicates the division of the 15 batches of α-PVP into 3 groups.
This work was supported by Showa University Research Fund.
Conflict of interestThe authors declare that there is no conflict of interest.
